Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi there! I’m Reginah Brian Nairobi, an AI Data Trainer and Evaluator based in Nairobi, Kenya. I’m passionate about optimizing large language models through precise data annotation, RLHF, and thoughtful prompt engineering, and I enjoy turning complex datasets into reliable, high-performing AI systems. With 4+ years of hands-on experience (since 2022), I’ve delivered 98%+ accuracy on challenging projects for Appen/CrowdGen and Sama. I thrive in collaborative, cross-functional environments, uphold data privacy, and continuously refine processes to reduce model hallucinations and biases while improving overall reliability.…Hi there! I’m Reginah Brian Nairobi, an AI Data Trainer and Evaluator based in Nairobi, Kenya. I’m passionate about optimizing large language models through precise data annotation, RLHF, and thoughtful prompt engineering, and I enjoy turning complex datasets into reliable, high-performing AI systems. With 4+ years of hands-on experience (since 2022), I’ve delivered 98%+ accuracy on challenging projects for Appen/CrowdGen and Sama. I thrive in collaborative, cross-functional environments, uphold data privacy, and continuously refine processes to reduce model hallucinations and biases while improving overall reliability.

Reginah Brian Nairobi

Data Scientist, AI Prompt Engineer, Audio Transcriber, +2





Hi there! I’m Reginah Brian Nairobi, an AI Data Trainer and Evaluator based in Nairobi, Kenya. I’m passionate about optimizing large language models through precise data annotation, RLHF, and thoughtful prompt engineering, and I enjoy turning complex datasets into reliable, high-performing AI systems. With 4+ years of hands-on experience (since 2022), I’ve delivered 98%+ accuracy on challenging projects for Appen/CrowdGen and Sama. I thrive in collaborative, cross-functional environments, uphold data privacy, and continuously refine processes to reduce model hallucinations and biases while improving overall reliability.…Hi there! I’m Reginah Brian Nairobi, an AI Data Trainer and Evaluator based in Nairobi, Kenya. I’m passionate about optimizing large language models through precise data annotation, RLHF, and thoughtful prompt engineering, and I enjoy turning complex datasets into reliable, high-performing AI systems. With 4+ years of hands-on experience (since 2022), I’ve delivered 98%+ accuracy on challenging projects for Appen/CrowdGen and Sama. I thrive in collaborative, cross-functional environments, uphold data privacy, and continuously refine processes to reduce model hallucinations and biases while improving overall reliability.

Available to hire

Hi there! I’m Reginah Brian Nairobi, an AI Data Trainer and Evaluator based in Nairobi, Kenya. I’m passionate about optimizing large language models through precise data annotation, RLHF, and thoughtful prompt engineering, and I enjoy turning complex datasets into reliable, high-performing AI systems.

With 4+ years of hands-on experience (since 2022), I’ve delivered 98%+ accuracy on challenging projects for Appen/CrowdGen and Sama. I thrive in collaborative, cross-functional environments, uphold data privacy, and continuously refine processes to reduce model hallucinations and biases while improving overall reliability.

Skills

AI Data Labelling

Experience Level

AI Data Labelling

Expert

Language

English

Fluent

Spanish; Castilian

Advanced

Swahili

Advanced

French

Advanced

Japanese

Beginner

Finnish

Beginner

Arabic

Intermediate

Work Experience

AI Data Trainer & Search Evaluator (Independent Contractor) at CrowdGen by Appen; OneForma; Uber AI

January 1, 2022 - Present

Prompt Engineering & Evaluation: Iteratively refined 500+ complex prompts to improve LLM response relevance, accuracy, and safety. Quality Assurance: Consistently achieved a 98% quality score in weekly audits for search engine evaluation and chatbot ranking. Data Labeling: Annotated large-scale datasets (Text/Image) for computer vision and NLP models, ensuring 100% adherence to project-specific guidelines. Error Analysis: Identified and flagged systematic model biases and hallucinations, contributing to a 15% improvement in model reliability.

Data Analyst / Content Moderator at Alligner

January 1, 2025 - March 31, 2026

Content Review: Evaluated user-generated content for policy compliance, maintaining a safe online environment for millions of users. Data Entry: Managed high-volume data migration with extreme precision, reducing entry errors by 20% through improved workflow templates.